| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.991111111111111, | |
| "eval_steps": 500, | |
| "global_step": 112, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017777777777777778, | |
| "grad_norm": 2.4493284225463867, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.8771, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.035555555555555556, | |
| "grad_norm": 2.5336005687713623, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.8888, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 2.3694822788238525, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8663, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07111111111111111, | |
| "grad_norm": 2.06482195854187, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.8508, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 1.5644984245300293, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.835, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 1.4116849899291992, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8097, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12444444444444444, | |
| "grad_norm": 1.1847808361053467, | |
| "learning_rate": 1.1666666666666668e-05, | |
| "loss": 0.7754, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 1.0902292728424072, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7364, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.21918785572052, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.7091, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.9935525059700012, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.6956, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19555555555555557, | |
| "grad_norm": 1.4380978345870972, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.6811, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 1.139083743095398, | |
| "learning_rate": 2e-05, | |
| "loss": 0.681, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2311111111111111, | |
| "grad_norm": 1.3757270574569702, | |
| "learning_rate": 1.9995065603657317e-05, | |
| "loss": 0.6716, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24888888888888888, | |
| "grad_norm": 1.0465571880340576, | |
| "learning_rate": 1.9980267284282718e-05, | |
| "loss": 0.635, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.9036633968353271, | |
| "learning_rate": 1.99556196460308e-05, | |
| "loss": 0.6447, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 0.8748593330383301, | |
| "learning_rate": 1.9921147013144782e-05, | |
| "loss": 0.6476, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3022222222222222, | |
| "grad_norm": 0.7611545920372009, | |
| "learning_rate": 1.9876883405951378e-05, | |
| "loss": 0.6097, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8435001373291016, | |
| "learning_rate": 1.982287250728689e-05, | |
| "loss": 0.6127, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3377777777777778, | |
| "grad_norm": 1.1226811408996582, | |
| "learning_rate": 1.9759167619387474e-05, | |
| "loss": 0.6112, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.6948593258857727, | |
| "learning_rate": 1.9685831611286312e-05, | |
| "loss": 0.6022, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 0.7424058318138123, | |
| "learning_rate": 1.9602936856769432e-05, | |
| "loss": 0.595, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.39111111111111113, | |
| "grad_norm": 0.6478083729743958, | |
| "learning_rate": 1.9510565162951538e-05, | |
| "loss": 0.6057, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.4088888888888889, | |
| "grad_norm": 0.6245326399803162, | |
| "learning_rate": 1.9408807689542257e-05, | |
| "loss": 0.6067, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 0.6616186499595642, | |
| "learning_rate": 1.9297764858882516e-05, | |
| "loss": 0.599, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.5162010192871094, | |
| "learning_rate": 1.9177546256839814e-05, | |
| "loss": 0.5856, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4622222222222222, | |
| "grad_norm": 0.46659624576568604, | |
| "learning_rate": 1.9048270524660197e-05, | |
| "loss": 0.5798, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.48762547969818115, | |
| "learning_rate": 1.891006524188368e-05, | |
| "loss": 0.5785, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.49777777777777776, | |
| "grad_norm": 0.425236314535141, | |
| "learning_rate": 1.8763066800438638e-05, | |
| "loss": 0.5885, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5155555555555555, | |
| "grad_norm": 0.4328088164329529, | |
| "learning_rate": 1.860742027003944e-05, | |
| "loss": 0.5897, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.495597779750824, | |
| "learning_rate": 1.8443279255020153e-05, | |
| "loss": 0.5739, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5511111111111111, | |
| "grad_norm": 0.3705892264842987, | |
| "learning_rate": 1.827080574274562e-05, | |
| "loss": 0.5742, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 0.35032135248184204, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 0.5548, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 0.3868001103401184, | |
| "learning_rate": 1.7901550123756906e-05, | |
| "loss": 0.5879, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6044444444444445, | |
| "grad_norm": 0.34849685430526733, | |
| "learning_rate": 1.7705132427757895e-05, | |
| "loss": 0.5588, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 0.3659916818141937, | |
| "learning_rate": 1.7501110696304598e-05, | |
| "loss": 0.5561, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.3479112982749939, | |
| "learning_rate": 1.7289686274214116e-05, | |
| "loss": 0.566, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6577777777777778, | |
| "grad_norm": 0.38614964485168457, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 0.5702, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6755555555555556, | |
| "grad_norm": 0.32522615790367126, | |
| "learning_rate": 1.684547105928689e-05, | |
| "loss": 0.5646, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 0.34715840220451355, | |
| "learning_rate": 1.661311865323652e-05, | |
| "loss": 0.5629, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.38434770703315735, | |
| "learning_rate": 1.63742398974869e-05, | |
| "loss": 0.5461, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7288888888888889, | |
| "grad_norm": 0.32962343096733093, | |
| "learning_rate": 1.6129070536529767e-05, | |
| "loss": 0.5668, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 0.37483224272727966, | |
| "learning_rate": 1.5877852522924733e-05, | |
| "loss": 0.5656, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7644444444444445, | |
| "grad_norm": 0.30097684264183044, | |
| "learning_rate": 1.5620833778521306e-05, | |
| "loss": 0.5529, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7822222222222223, | |
| "grad_norm": 0.33267274498939514, | |
| "learning_rate": 1.5358267949789968e-05, | |
| "loss": 0.5563, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.33388659358024597, | |
| "learning_rate": 1.5090414157503715e-05, | |
| "loss": 0.5665, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8177777777777778, | |
| "grad_norm": 0.3299979269504547, | |
| "learning_rate": 1.4817536741017153e-05, | |
| "loss": 0.5527, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8355555555555556, | |
| "grad_norm": 0.30535200238227844, | |
| "learning_rate": 1.4539904997395468e-05, | |
| "loss": 0.5662, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 0.32509496808052063, | |
| "learning_rate": 1.4257792915650728e-05, | |
| "loss": 0.5447, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8711111111111111, | |
| "grad_norm": 0.3176334798336029, | |
| "learning_rate": 1.3971478906347806e-05, | |
| "loss": 0.5464, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.30327945947647095, | |
| "learning_rate": 1.3681245526846782e-05, | |
| "loss": 0.5449, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 0.2697887718677521, | |
| "learning_rate": 1.3387379202452917e-05, | |
| "loss": 0.54, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9244444444444444, | |
| "grad_norm": 0.2823696434497833, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 0.5543, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9422222222222222, | |
| "grad_norm": 0.30424606800079346, | |
| "learning_rate": 1.2789911060392295e-05, | |
| "loss": 0.5514, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.26427096128463745, | |
| "learning_rate": 1.2486898871648552e-05, | |
| "loss": 0.5588, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 0.27375537157058716, | |
| "learning_rate": 1.2181432413965428e-05, | |
| "loss": 0.5429, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9955555555555555, | |
| "grad_norm": 0.2701137363910675, | |
| "learning_rate": 1.187381314585725e-05, | |
| "loss": 0.5585, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0133333333333334, | |
| "grad_norm": 0.5673031806945801, | |
| "learning_rate": 1.156434465040231e-05, | |
| "loss": 0.9233, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.031111111111111, | |
| "grad_norm": 0.29599809646606445, | |
| "learning_rate": 1.1253332335643043e-05, | |
| "loss": 0.544, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.048888888888889, | |
| "grad_norm": 0.3540167808532715, | |
| "learning_rate": 1.0941083133185146e-05, | |
| "loss": 0.5149, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.3205398917198181, | |
| "learning_rate": 1.0627905195293135e-05, | |
| "loss": 0.4813, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0844444444444445, | |
| "grad_norm": 0.35224997997283936, | |
| "learning_rate": 1.0314107590781284e-05, | |
| "loss": 0.5588, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.1022222222222222, | |
| "grad_norm": 0.2714930474758148, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4924, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.2932974398136139, | |
| "learning_rate": 9.685892409218718e-06, | |
| "loss": 0.4994, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1377777777777778, | |
| "grad_norm": 0.29003581404685974, | |
| "learning_rate": 9.372094804706867e-06, | |
| "loss": 0.4862, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 0.31004598736763, | |
| "learning_rate": 9.058916866814857e-06, | |
| "loss": 0.5294, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1733333333333333, | |
| "grad_norm": 0.26587504148483276, | |
| "learning_rate": 8.746667664356957e-06, | |
| "loss": 0.5222, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1911111111111112, | |
| "grad_norm": 0.25245964527130127, | |
| "learning_rate": 8.43565534959769e-06, | |
| "loss": 0.4697, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.208888888888889, | |
| "grad_norm": 0.27027666568756104, | |
| "learning_rate": 8.126186854142752e-06, | |
| "loss": 0.5343, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2266666666666666, | |
| "grad_norm": 0.25415799021720886, | |
| "learning_rate": 7.818567586034578e-06, | |
| "loss": 0.4861, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 0.25560981035232544, | |
| "learning_rate": 7.513101128351454e-06, | |
| "loss": 0.5141, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2622222222222224, | |
| "grad_norm": 0.2490381896495819, | |
| "learning_rate": 7.210088939607709e-06, | |
| "loss": 0.4685, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.2686901092529297, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 0.5433, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.2977777777777777, | |
| "grad_norm": 0.19526812434196472, | |
| "learning_rate": 6.612620797547087e-06, | |
| "loss": 0.4569, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.3155555555555556, | |
| "grad_norm": 0.2415803074836731, | |
| "learning_rate": 6.318754473153221e-06, | |
| "loss": 0.5253, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.19613640010356903, | |
| "learning_rate": 6.028521093652195e-06, | |
| "loss": 0.4791, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3511111111111112, | |
| "grad_norm": 0.23172248899936676, | |
| "learning_rate": 5.742207084349274e-06, | |
| "loss": 0.5268, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3688888888888888, | |
| "grad_norm": 0.22319753468036652, | |
| "learning_rate": 5.460095002604533e-06, | |
| "loss": 0.5033, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.3866666666666667, | |
| "grad_norm": 0.253582626581192, | |
| "learning_rate": 5.1824632589828465e-06, | |
| "loss": 0.5263, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.4044444444444444, | |
| "grad_norm": 0.22795237600803375, | |
| "learning_rate": 4.909585842496287e-06, | |
| "loss": 0.4646, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.21554000675678253, | |
| "learning_rate": 4.641732050210032e-06, | |
| "loss": 0.4976, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.22318841516971588, | |
| "learning_rate": 4.379166221478697e-06, | |
| "loss": 0.5116, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4577777777777778, | |
| "grad_norm": 0.22658716142177582, | |
| "learning_rate": 4.12214747707527e-06, | |
| "loss": 0.5323, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4755555555555555, | |
| "grad_norm": 0.20563125610351562, | |
| "learning_rate": 3.8709294634702374e-06, | |
| "loss": 0.4499, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4933333333333334, | |
| "grad_norm": 0.20641934871673584, | |
| "learning_rate": 3.625760102513103e-06, | |
| "loss": 0.4826, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 0.2252846360206604, | |
| "learning_rate": 3.3868813467634833e-06, | |
| "loss": 0.5218, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.528888888888889, | |
| "grad_norm": 0.24224503338336945, | |
| "learning_rate": 3.1545289407131128e-06, | |
| "loss": 0.4953, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.5466666666666666, | |
| "grad_norm": 0.20531503856182098, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.5141, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5644444444444443, | |
| "grad_norm": 0.2080065906047821, | |
| "learning_rate": 2.7103137257858867e-06, | |
| "loss": 0.5073, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5822222222222222, | |
| "grad_norm": 0.19242417812347412, | |
| "learning_rate": 2.4988893036954045e-06, | |
| "loss": 0.4792, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.2007436901330948, | |
| "learning_rate": 2.2948675722421086e-06, | |
| "loss": 0.5097, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6177777777777778, | |
| "grad_norm": 0.19588816165924072, | |
| "learning_rate": 2.098449876243096e-06, | |
| "loss": 0.4919, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6355555555555554, | |
| "grad_norm": 0.19348369538784027, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.5275, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.6533333333333333, | |
| "grad_norm": 0.19428110122680664, | |
| "learning_rate": 1.7291942572543806e-06, | |
| "loss": 0.4784, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6711111111111112, | |
| "grad_norm": 0.1935623586177826, | |
| "learning_rate": 1.5567207449798517e-06, | |
| "loss": 0.5102, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 0.20146863162517548, | |
| "learning_rate": 1.3925797299605649e-06, | |
| "loss": 0.4932, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.7066666666666666, | |
| "grad_norm": 0.2071317434310913, | |
| "learning_rate": 1.2369331995613664e-06, | |
| "loss": 0.5106, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.7244444444444444, | |
| "grad_norm": 0.1940830498933792, | |
| "learning_rate": 1.0899347581163222e-06, | |
| "loss": 0.51, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.7422222222222223, | |
| "grad_norm": 0.18540237843990326, | |
| "learning_rate": 9.517294753398066e-07, | |
| "loss": 0.4936, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.18730789422988892, | |
| "learning_rate": 8.224537431601886e-07, | |
| "loss": 0.4605, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.18820004165172577, | |
| "learning_rate": 7.022351411174866e-07, | |
| "loss": 0.5104, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7955555555555556, | |
| "grad_norm": 0.18577077984809875, | |
| "learning_rate": 5.911923104577455e-07, | |
| "loss": 0.4996, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.8133333333333335, | |
| "grad_norm": 0.19338218867778778, | |
| "learning_rate": 4.894348370484648e-07, | |
| "loss": 0.4835, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.8311111111111111, | |
| "grad_norm": 0.19661261141300201, | |
| "learning_rate": 3.9706314323056936e-07, | |
| "loss": 0.4947, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8488888888888888, | |
| "grad_norm": 0.19002115726470947, | |
| "learning_rate": 3.1416838871368925e-07, | |
| "loss": 0.4762, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 0.18881858885288239, | |
| "learning_rate": 2.4083238061252565e-07, | |
| "loss": 0.5344, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8844444444444446, | |
| "grad_norm": 0.18790070712566376, | |
| "learning_rate": 1.7712749271311392e-07, | |
| "loss": 0.4645, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.9022222222222223, | |
| "grad_norm": 0.19742755591869354, | |
| "learning_rate": 1.231165940486234e-07, | |
| "loss": 0.4809, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.19038796424865723, | |
| "learning_rate": 7.885298685522235e-08, | |
| "loss": 0.4858, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.9377777777777778, | |
| "grad_norm": 0.1900634616613388, | |
| "learning_rate": 4.438035396920004e-08, | |
| "loss": 0.5078, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 0.19468751549720764, | |
| "learning_rate": 1.973271571728441e-08, | |
| "loss": 0.5179, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9733333333333334, | |
| "grad_norm": 0.1910133957862854, | |
| "learning_rate": 4.9343963426840006e-09, | |
| "loss": 0.4961, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.991111111111111, | |
| "grad_norm": 0.19184480607509613, | |
| "learning_rate": 0.0, | |
| "loss": 0.511, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.991111111111111, | |
| "step": 112, | |
| "total_flos": 621868991119360.0, | |
| "train_loss": 0.5655743213636535, | |
| "train_runtime": 48561.3008, | |
| "train_samples_per_second": 0.222, | |
| "train_steps_per_second": 0.002 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 112, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 621868991119360.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |